One for all, all for one: Agents with social identities

One of the challenges in developing multi-agent systems is the creation of agents able to exhibit human-like behaviours in complex social situations. In order to do so, agents need to be socially aware of their environment and perceive other agents not only as individuals but also as social group members. Following Social Identity and Self-Categorization theories, we developed the Dynamic Identity Model for Agents that provides agents with the ability to adapt their identity and behaviour to the social context. We then implemented it in a social dilemma scenario where different situations were explored.